The purpose of this proposal is to respond to the Request for Applications for cooperative agreement RFA number 90-CA-05, National Collaborative Diagnostic Imaging Trial Projects, National Cancer Institute. Appropriate staging of a malignancy is a crucial guide to patient prognosis and management. This is especially true for musculoskeletal sarcomas. Knowledge of lesional extent helps to both determine the appropriate surgical procedure and define appropriate surgical margins for adequate treatment. Several investigators have compared MRI and CT in their ability to detect local lesional extent in musculoskeletal sarcomas. Comparison of these studies is, however, difficult due to the broad range of experimental designs and protocols used. Recent investigations have assessed the ability of MRI to evaluate the response to musculoskeletal sarcomas to preoperative chemotherapy. Again, comparison of these studies is difficult, and it is still unknown when these modalities should be employed, and which to use. The current cost for imaging studies during the first year following diagnosis of a musculoskeletal sarcoma is between $14,566 and $18,652. A portion of this expense is attributable to the fact that often, both MRI and CT scans are obtained when one would suffice. It is hoped that information obtained from this study will assist in establishing imaging protocols which will yield maximum diagnostic information while eliminating duplicity of examinations and reduce the total health care cost. The aim of this study is to assess the ability of magnetic resonance imaging (with and without IV contrast enhancement) and computed tomography (with IV contract enhancement) to stage and determine local extent of biopsy proven musculoskeletal sarcomas prior to therapeutic intervention, following preoperative chemotherapy (where applicable), and following surgical intervention. Tumors to be included are liposarcoma, soft tissue malignant fibrous histiocytoma, synovial sarcoma, soft tissue fibrosarcoma, neurosarcoma, osteosarcoma, and chondrosarcoma. It is estimated that 50 patients per year will be enrolled in this study at UCLA. Imaging studies will be concerned with evaluation of the primary tumor prior to surgical intervention and monitoring for local tumor recurrence following surgery. Radiographs, MR scans, and CT scans will be analyzed by four Musculoskeletal Radiologists. Clinical readings will be done separately from research readings. Research readings will be accomplished both without and with clinical information, and the responsibility for reading the different modalities within each set of images will be rotated. Specific features which will be addressed in the readings include: 1) maximum length and width of tumor, 2) compartmentalization, 3) fibrous capsule/pseudocapsule, 4) peritumoral edema/ inflammation, 5) relationship to major neurovascular bundle, and 6) relationship to nearest joint. Determination of tumor type will be obtained by histologic evaluation of tissue specimens obtained at open incisional biopsy. Following definitive surgery, specimens will be analyzed in order to determine the sensitivity of the readers to determine tumor stage and local extent. This will include whole section analysis and histologic study of tumor margins.